About

Nicholas Hopper is a Professor and Associate Head for Instruction in the Computer Science & Engineering Department at the University of Minnesota. His primary research interest lies in online communications security and privacy, with a main focus on anonymous communication, censorship resistance, and applied cryptography. Over the course of his career, he has also worked on security and privacy issues in peer-to-peer networks, BGP routing, electronic cash, and provable security for steganography and watermarking.

Research topics

• Computer Science
• Artificial Intelligence
• Computer Security
• Machine Learning
• Chemical engineering
• Chemistry
• Physics
• World Wide Web
• Engineering
• Classical mechanics
• Programming language
• Materials science

Selected publications

• Tracing the Chain: Deep Learning for Stepping-Stone Intrusion Detection

  ArXiv.org · 2026-04-09

  articleOpen access

  Stepping-stone intrusions (SSIs) are a prevalent network evasion technique in which attackers route sessions through chains of compromised intermediate hosts to obscure their origin. Effective SSI detection requires correlating the incoming and outgoing flows at each relay host at extremely low false positive rates -- a stringent requirement that renders classical statistical methods inadequate in operational settings. We apply ESPRESSO, a deep learning flow correlation model combining a transformer-based feature extraction network, time-aligned multi-channel interval features, and online triplet metric learning, to the problem of stepping-stone intrusion detection. To support training and evaluation, we develop a synthetic data collection tool that generates realistic stepping-stone traffic across five tunneling protocols: SSH, SOCAT, ICMP, DNS, and mixed multi-protocol chains. Across all five protocols and in both host-mode and network-mode detection scenarios, ESPRESSO substantially outperforms the state-of-the-art DeepCoFFEA baseline, achieving a true positive rate exceeding 0.99 at a false positive rate of $10^{-3}$ for standard bursty protocols in network-mode. We further demonstrate chain length prediction as a tool for distinguishing malicious from benign pivoting, and conduct a systematic robustness analysis revealing that timing-based perturbations are the primary vulnerability of correlation-based stepping-stone detectors.

  PublisherOA PDF

• Tracing the Chain: Deep Learning for Stepping-Stone Intrusion Detection

  arXiv (Cornell University) · 2026-04-09

  preprintOpen access

  Stepping-stone intrusions (SSIs) are a prevalent network evasion technique in which attackers route sessions through chains of compromised intermediate hosts to obscure their origin. Effective SSI detection requires correlating the incoming and outgoing flows at each relay host at extremely low false positive rates -- a stringent requirement that renders classical statistical methods inadequate in operational settings. We apply ESPRESSO, a deep learning flow correlation model combining a transformer-based feature extraction network, time-aligned multi-channel interval features, and online triplet metric learning, to the problem of stepping-stone intrusion detection. To support training and evaluation, we develop a synthetic data collection tool that generates realistic stepping-stone traffic across five tunneling protocols: SSH, SOCAT, ICMP, DNS, and mixed multi-protocol chains. Across all five protocols and in both host-mode and network-mode detection scenarios, ESPRESSO substantially outperforms the state-of-the-art DeepCoFFEA baseline, achieving a true positive rate exceeding 0.99 at a false positive rate of $10^{-3}$ for standard bursty protocols in network-mode. We further demonstrate chain length prediction as a tool for distinguishing malicious from benign pivoting, and conduct a systematic robustness analysis revealing that timing-based perturbations are the primary vulnerability of correlation-based stepping-stone detectors.

  PublisherDOI

• No Safety in Numbers: Traffic Analysis of Sealed-Sender Groups in Signal

  2025-08-26

  articleSenior author

  Signal messenger is a popular server-mediated end-to-end encrypted messaging application, and its underlying encryption protocol provides many attractive properties such as authentication, post-compromise security and deniability. Signal additionally provides two extensions, Sealed Sender and Private Groups, to conceal metadata about who communicates with whom, and which parties communicate in groups, from potentially compromised Signal servers. We describe a novel attack for group conversations in Signal and show through theoretical analysis and simulation that groups of communicating entities may be linked through recipient metadata alone, defeating the Sealed Sender and Private Groups mechanisms. We show that an earlier defense proposed for a similar attack on Sealed Sender pairs is not effective against our attack. We then discuss a “server-agnostic” mitigation that can be implemented by users alone, illustrating a tradeoff between communication overhead and defense efficacy.

  PublisherDOI

• Shear Thinning and Stress-Dependent Viscosity Activation Volumes: Combining Eyring and Carreau

  Tribology Letters · 2025-07-23 · 1 citations

  articleOpen access1st author

  Abstract The viscosity of fluids and their dependence on shear rate, known as shear thinning, plays a critical role in applications ranging from lubricants and coatings to biomedical and food-processing industries. Traditional models such as the Carreau and Eyring theories offer competing explanations for shear-thinning behavior. The Carreau model attributes viscosity reduction to molecular distortions, while the Eyring model describes shear thinning as a stress-induced transition over an activation energy barrier. This work proposes an extended-Eyring model that incorporates stress-dependent activation volumes, bridging key aspects of both theories. In modifying transition-state theory by using an Evans-Polanyi perturbation analysis, we derive a generalized viscosity equation that accounts for the molecular-scale rearrangements governing fluid flow. The model is validated against computational and experimental data, including shear-thinning behavior of pure squalane and polyethylene oxide (PEO) aqueous solutions. Comparative analysis with Carreau-Yasuda and conventional Eyring models demonstrates excellent accuracy in predicting viscosity trends over a wide range of shear rates. The introduction of stress-dependent activation volumes provides a description of molecular exchange kinetics accounting for structural reorganization under shear. These findings offer a unified framework for modeling shear thinning and have broad implications for designing advanced lubricants, polymer solutions, and complex fluids with tailored flow properties. Graphical Abstract

  PublisherOA PDFDOI

• Abstract 1579 Analogs of NIH Molecular Probe ML283 Identify a Possible Druggable Pocket on the SARS-CoV-2 nsp13 Helicase

  Journal of Biological Chemistry · 2025-05-01

  articleOpen access

  RNA virus-encoded helicases are essential for viral replication yet with sufficiently similar structures that antivirals targeting one might work well for many other pathogens with pandemic potential. This study examines the effects of National Institutes of Health molecular probe ML283 and 53 related compounds on the SARS-CoV-2 nsp13 helicase. ML283 was synthesized as a potent, selective inhibitor of the helicase encoded by the hepatitis C virus. Unlike prior work with the HCV and Dengue virus helicase, there appears to be an apparent disconnect between the ability of the ML283 analogs to inhibit both unwinding and helicase-catalyzed ATP hydrolysis.

  PublisherOA PDFDOI

• Activation Volumes in Tribochemistry; What Do They Mean and How to Calculate Them?

  Tribology Letters · 2025-02-24 · 5 citations

  articleOpen access1st authorCorresponding
  PublisherOA PDFDOI

• On the pressure dependence of viscosity, especially for fluids that have a tendency to form glasses

  The Journal of Chemical Physics · 2024-12-04 · 2 citations

  article1st authorCorresponding

  Understanding fluid viscosity is crucial for applications including lubrication and chemical kinetics. A commonality of molecular models is that they describe fluid flow based on the availability of vacant space. The proposed analysis builds on Goldstein's idea that viscous transport must involve the concerted motion of a molecular ensemble, referred to as cooperatively rearranging regions (CRRs) by Adam and Gibbs in their entropy-based viscosity model for liquids close to their glass transition. The viscosity data for propylene carbonate reveal a non-monotonic trend of the activation volume with pressure, suggesting the existence of two types of CRR with different compressibility behaviors. This is proposed to result from a change in CRR free volume (<0.2 GPa) and a growth in its size (>0.2 GPa). We use Evans-Polanyi perturbation theory to develop an analytical model for the structural changes of the CRR in function of pressure and temperature and their effect on Eyring viscosity. This analysis shows that the activation energies and volumes scale with the CRR size. Using the compressibility data of propylene carbonate, we show that the activation volume of the CRR at low pressures depends on the compressibility of an ensemble comprised of the first coordination shell around a molecule. At higher pressures, we apply an Adam-Gibbs-type analysis to model the increase in CRR size and its effect on viscosity, where the increase in size is estimated from propylene carbonate's heat capacity. However, this analysis also reveals deviations from the Adam and Gibbs model that will guide future improvements.

  PublisherDOI

• An Analysis of Shear-Dependent Mechanochemical Reaction Kinetics

  Research Square · 2024-02-12

  preprintOpen access

  <title>Abstract</title> The variation in the rate of a tribochemical reaction is calculated as a function of combined normal and shear stresses using Evans-Polanyi perturbation theory. The effect of perturbations such as stresses is obtained using transition-state theory from their influence on the equilibrium constant between the initial- and transition-state structures using the molar Gibbs free energy change. An advantage of this approach is it capability of calculating the effect of several perturbations, such as combined normal and shear stresses. Two effects have been identified. The first is that the effective activation volume contains contributions from both the normal and shear stresses. More importantly, the analysis predicts that the asymptote of this plot at zero stress is not equal to the thermal reaction rate; there is a change in the inherent tribochemical reaction rate that depends on velocity. This prediction is shown to be true for the shear-induced decomposition of ethyl thiolate species adsorbed on a Cu(100) single crystal substrate where this effect contributes to about two orders of magnitude increase in the reaction rate. This indicates that tribochemical reactions can be influenced by either just normal stresses or a combination of normal and shear stresses, but that the latter contribution is much larger. It is predicted that there is a linear relationship between the activation energy and the logarithm of the pr-exponential factor of this asymptotic rate constant, known as a compensation effect in catalysis. While this has not yet been seen for tribochemical reactions on surfaces, it has been found for reactions occurring in sheared fluids.

  PublisherOA PDFDOI

• Modeling mechanochemistry: pressure dependence of Diels–Alder cycloaddition reaction kinetics

  RSC Mechanochemistry · 2024-01-01 · 9 citations

  articleOpen access1st author

  We analyze the effect of pressure on the Diels–Alder (D–A) dimerization reactions using Evans–Polanyi (E–P) theory, a thermodynamic analysis of the way in which a perturbation, in this case a hydrostatic pressure, modifies a reaction rate.

  PublisherOA PDFDOI

• An Analysis of Shear-Dependent Mechanochemical Reaction Kinetics

  Tribology Letters · 2024-06-12 · 7 citations

  articleOpen access
  PublisherOA PDFDOI

Recent grants

• CT-ISG: A Framework for Trustworthy Cooperative P2P Applications

  NSF · $222k · 2007–2011

• CAREER: Practice-oriented provably secure information hiding

  NSF · $422k · 2006–2012

• SaTC: CORE: Small: Models and Measurements for Website Fingerprinting

  NSF · $500k · 2018–2023

• TWC: Option: Medium: Measurement-Based Design and Analysis of Censorship Circumvention Schemes

  NSF · $1.2M · 2013–2018

• SaTC: CORE: Small: Consistent and Private Group Communication

  NSF · $500k · 2018–2023

Frequent coauthors

• Yongdae Kim

  47 shared

• Wilfred T. Tysoe

  University of Wisconsin–Milwaukee

  32 shared

• François Sidoroff

  Laboratoire de Tribologie et Dynamique des Systèmes

  26 shared

• Resham Rana

  University of Wisconsin–Milwaukee

  21 shared

• Eric Chan‐Tin

  Loyola University Chicago

  16 shared

• Eugene Y. Vasserman

  15 shared

• Juliette Cayer-Barrioz

  Laboratoire de Tribologie et Dynamique des Systèmes

  14 shared

• Denis Mazuyer

  École Centrale de Lyon

  14 shared

Awards & honors

• McKnight Land-Grant Professorship (2008)
• Science and Engineering Student Board - Best Professor Award…
• National Science Foundation Faculty Early Career Development…